This study constructs spillover indices from a volatility spillover network perspective using the Quantile Vector Autoregression (QVAR) model, capturing the RMB exchange rate's tail risk and time-frequency effects across varying shock sizes. Empirical results show that the QVAR-based spillover index more effectively captures the tail risk spillover effects across different quantiles. In the time domain, spillovers between RMB exchange rates are dynamic and particularly sensitive to extreme contingencies. In the frequency domain, RMB exchange rates demonstrate significant spillovers, primarily at low frequencies. During extreme upward events, dynamic observations show high-frequency spillovers surpassing low-frequency ones as dominant drivers in the tail spillovers of the RMB exchange rate. Additionally, the analysis of tail dependence indicators indicates a strong asymmetry in RMB exchange rate correlations, emphasizing market participants' heightened sensitivity to unfavorable shocks. These findings can serve as a reference for policymakers to strengthen risk management of the RMB exchange rate.
Read full abstract